Selective oxygenation of aliphatic ethers catalyzed by ruthenium(II

Selective oxygenation of aliphatic ethers catalyzed by ruthenium(II) complexes. Mario Bressan, Antonino Morvillo, and Giorgio Romanello. Inorg. Chem. ...
0 downloads 0 Views 533KB Size
2976

Inorg. Chem. 1990, 29, 2976-2979 Contribution from the Dipartimento di Chimica Inorganica, Universitl di Padova, via Marzolo 1, 35 131 Padova, Italy

Selective Oxygenation of Aliphatic Ethers Catalyzed by Ruthenium(I1) Complexes Mario Bressan,* Antonino Morvillo, and Giorgio Romanello Received October 25, 1989 Selective and fast conversions of primary ethers to esters (or lactones) are achieved by using lithium hypochlorite in the presence of catalytic amounts of complexes of ruthenium(I1) RuX2L4[L = (CH,)sO, (L), = 1,3-bis(diphenylphosphino)propane]. Isopropyl ether undergoes oxidative cleavage to acetone, while secondary ethers bearing methylene groups in the &position are converted into @-ketoethers. Methyl ethers are unaffected by the oxidation. The origin of the observed selectivity is investigated and discussed.

Introduction Metal oxides in high oxidation states, such as chromium trioxide and permanganate, are among the few stoichiometric oxidants that are effective for the selective oxygenation of primary ethers to esters.la*b Reactions with other conventional oxidizing agents, such as bromine, lead tetraacetate, or ozone," as well as anodic oxidation," lead to further oxidation of the esters to anhydrides and carboxylic acids. Ruthenium tetraoxide, early introduced as a useful stoichiometric reagent for the mild and selective oxidation of a variety of organic substrates, also promotes the oxidation of primary ethers to esters with moderate selectivity.2a Attempts to make the above oxidation catalytic by using simple salts of ruthenium in conjugation with periodate or hypochlorite ions as the stoichiometric oxidants are also Other mild catalytic methods, based on heterogeneous platinum catalysts and dioxygen, which are able to unselectively convert cyclic ethers into lactones and products of oxidative cleavage are d e ~ c r i b e d . ~ Following our previous communication,4 dealing with the catalytic behavior of a representative choice of ruthenium(I1) complexes in the oxygen-transfer reactions from single-oxygen donors to organic substrates, we delineate here some mechanistic features for the oxidation of ethers. The catalysts examined are five- and six-coordinate complexes of ruthenium(II), namely [RUCI(DPP)~]PF~, r r ~ n s - [ R u C l ~ ( D P P )[DPP ~ ] = 1,3-bis(diphenyl phosphino) propane], and cis-(RuCI2[(CH3),SO] 4). cis[RuCI2(1,IO-phenanthr~line)~] and the osmium(I1) derivative [OSCI(DPP)~]PF~ were occasionally tested but exhibited lower activity. Lithium hypochlorite or sodium hypochlorite, which was successfully introduced for the oxidation of organic substrates catalyzed by metalloporphyrin~,~ and transition-metal salts6 were used as alkaline aqueous solutions. In the presence of catalytic amounts of the ruthenium(I1) complexes, hypochlorite exhibits good oxidizing properties, thus possibly making the present procedure a fast, mild, and chemioselective method for the oxygenation of ethers. Experimental Section Materials. [RuCI(DPP),]PF, was prepared by refluxing ethanolic suspensions of rrans-[RuCI2(DPP),]in the presence of excess NH4PFp,' and cis-{RuC12[(CH,),SO],)by the standard procedure.8 Commercial products were used for gas chromatographic standardization. Tetrahydrofuran-d8 (98%) was purchased from Aldrich, while tetrahydrofuran-3,3,4,4-d4was prepared by acid-catalyzed (Nation-H) condensation of 98% I ,4-butanediol-2.2,3,3-d4, following the procedure described by (a) Harrison, 1. T.;Harrison, S. J . Cfiem. Soc., Cfiem. Commun. 1966, 752. (b) Schmidt, H.J.; Schaefer, H. J. Angew. Cfiem., Inr. Ed. Engl. 1979, 18.69. (c) Deno, N. C.; Potter, N. H. J . Am. Cfiem. SOC.1967, 89, 3550. (d) Franze, V.;Edens, R. Justus Liebigs Ann. Cfiem. 1970, 735, 47. (e) Pollart, K. A.; Mueller, R. J . Org. Cfiem. 1962,27,2392. (0 Shono, T.;Matsumura, Y. J . Am. Cfiem. Soc. 1969, 91, 2803. (a) Berkowitz, L. M.; Rylander, P. N. J . Am. Cfiem. SOC.1958, 80, 6682; US. Patent 3,278,558;Cfiem. Absrr. 1967, 66. 105978. (b) Smith, A. B.; Scarborough, R. M. Synrfi. Commun. 1980,IO, 205. (c) Carlsen, P. H. J.; Katsuki, T.; Martin, V. S.; Sharpless, K. B. J . Org. Cfiem. 1981, 46. 3336. Heyns, K.; Buchholz, H. Cfiem. Eer. 1976, 109, 3707. Bressan. M.; Morvillo. A. J . Cfiem. Soc.. Cfiem. Commun. 1989, 421. Meunier, B. Bull. SOC.Cfiim. Fr. 1986, 578. Mijs, W. J.; deJonge, C . R. H. I. Organic Syntheses by Oxidation with Metal Compounds;Plenum Press: New York, 1986. Bressan. M.; Rigo, P. Inorg. Chem. 1975, 14, 2286. Evans. 1. P.; Spencer, A.; Wilkinson. G.J . Cfiem. Soc., Dalton Trans. 1973. 204.

0020- 1669/90/ 1329-2976$02.50/0

Table I. Ruthenium-Catalvzed Oxidation of Ethers bv LiCIo" ~

~~

substrate di-n-propyl ether diisopropyl ether methyl n-butyl

ether tetrahydrofuran 2-methyltetrahydrofuran 2,5-dimethyltetrahydrofuran 2,5-dimethylfuran 2,3-dihydrofuran

catalytic activityb I I1 111

products propyl propionate

51

acetone methyl butyrate

c 24

c

y-butyrolactone 2-meth yldihydrofuran-3-one

42 12

54 21 12 6

y-valerolactone

9

6

66 24 54 24 15

6

2,5-dimethyldihydrofuran-3-one 177 93 52

2,5-dimethylfuran-3-one y-butyrolactone

c c

c

4.4

c

0.1

OCatalyst, 2 mM, and substrate, 1 M, in CH2C12;LiCIO, 0.7 M, in H20;22 OC. turnovers/h. I = c ~ ~ - ( R u C I ~ [ ( C H ~ )I1~ S=O ] ~ ] ; trans-[RuCl,(DPP),]; 111 = [RuCI(DPP),]PF,. CNotdetermined. Olah.9 Isotopic purity of the product was checked by 'H NMR and mass spectroscopy. Instrumentation. NMR spectra were recorded on a JEOL FX 90 Q FT spectrometer. Organic analyses were made on a DAN1 3800 gas chromatograph equipped with FID, using a 30-m SE-30 capillary column (0.25-mm i.d.). The identity of each product was confirmed by comparison of the fragmentation patterns in the mass spectra obtained with a VG 16F mass spectrometer operating in the electron ionization mode at 70 eV. Procedure for the Catalytic Oxidations. A I-mL aliquot of a CH2C12 solution containing 1 mmol of the substrates, 2 rmol of the complexes, and a gas chromatographic standard (n-decane, 5 rmol) was stirred magnetically in a 2-mL vial together with 1 mL of an aqueous alkaline solution of lithium hypochlorite (between 0.5 and 1 M, as determined by iodometric titration). GLC analyses were performed on aliquots of the organic layer, withdrawn at appropriate times with a hypodermic microsyringe. Reaction rates were determined at 22 OC for early stages of the reaction (up to 10 turnovers) and were reproducible to within 15%. Results The catalytic oxygenation of a choice of secondary and primary ethers is investigated. In a typical experiment, the ethereal substrates (1 M) and the catalyst (2 mM) are dissolved in CH2Cl2 and the resulting solution is stirred vigorously with an equal amount of an aqueous alkaline (pH ca. 13) solution of LiCIO, 0.7 M. Data reported in Table I refer to the initial rate of the conversions as measured from the amount of the products detected in the organic layer. Kinetic isotope effects were measured for the oxygenation of variously deuterated forms of tetrahydrofuran into the corresponding y-butyrolactones, and the results are presented in Table 11. A kinetic investigation was undertaken for the ruthenium(I1)-catalyzed reaction between 2,5-dimethyltetrahydrofuran and hypochlorite, whose results are presented in Figure 2. The reaction mechanism of Scheme I1 is proposed, which will be discussed at appropriate places in the following section. Discussion The compounds listed in Table I are the only products detected in the reaction mixtures, apart from negligible amounts of chlorinated derivatives (ca. 1% or less), likely arising from the chlo( 9 ) Olah, G.A.; Fung, A.

P.; Malhotra, R. Synthesis 1981, 474.

0 . 1990 American Chemical Societv

Inorganic Chemistry, Vol. 29, No. 16, 1990 2977

Selective Oxygenation of Aliphatic Ethers Scheme I

0 0

40

20

60

80

100

time, h -~

Figure 1. Time course of the oxygenation of methyl n-butyl ether (0) to methyl butyrate (m) by aqueous LiClO (?#)catalyzed by [RuCI(DPP)JPF,. 2 mM, in CH2CI2at 22 OC. + refers to the concentration of residual LiClO in an experiment of dismutation in the same conditions and in the absence of the substrate. Arrows indicate the addition of fresh solid LiCIO.

rinating action of small amounts of HCIO: they were not formed when other single-oxygen oxidants, such as PhIO or KHS05, were used. Rigorous degassing of the solutions, to remove oxygen prior to the reactions, has no effect on both the composition and the rate of formation of the products. Conversions may be somewhat larger than those reported in Table I, since some oxidation products are moderately soluble in the aqueous saline phase. The latter, previously acidified with H2S04,was carefully examined to check for the presence of carboxylic and hydroxy carboxylic acids, which are among the possible products of the oxidation or the alkaline hydrolysis of the formed esters and lactones. In fact, although some of the esters and lactones produced by the reactions under investigation are reported to undergo fairly rapid hydrolysis in alkaline media,I0 we never detected significant amounts of hydrolysis products within the short reaction times examined. The P I esent catalytic system generally maintains the initial activities for an extended period of time, and over 1000 cycles have been achieved, provided that an excess of hypochlorite, calculated on the basis of the stoichiometric requirement, is periodically added to the reaction mixtures (Figure l). The losses of oxidant can be explained by noting that the ruthenium complexes also catalyze the dismutation of hypochlorite to chloride and oxygen, which is a well-known process, commonly triggered by a variety of metal ions." Test experiments showed that 2 mM CH2CI2solutions of [RuCI(DPP),]PF, promoted the dismutation of 1 M aqueous solutions of LiClO at an initial rate of ca. 30 turnovers/h. Dismutation is favored by the presence of small amounts of ethers, less than 0.1 M, which likely act as phasetransfer agents with respect to hypochlorite: indeed, addition of conventional phase-transfer agents, i.e. crown ethers and quaternary ammonium salts, does not significantly improve yields and conversion rates. Analysis of the oxidation mixtures of primary ethers RCH20CH2Rrevealed the presence of esters (or lactones) only (eq 1). No products of further oxidation, such as anhydrides (eq 2) or carboxylic acids, derived from their hydrolysis, were detected. RCH2OCHZR 2 [ 0 ] RCH2OCOR + H2O (1) RCHZOCOR + 2 [ 0 ] RCOOCOR + H 2 0 (2)

+

-

+

We independently verified that esters and lactones are recovered unchanged, if reacted with the ruthenium(I1)-hypochlorite system under identical conditions. It is noteworthy that traditional ruthenium tetraoxide oxidations are, on the contrary, rather unselective, giving rise to extensive (IO) Blackburn, G.M.; Dodds, H.L. M.J . Chem. SOC.,Perkin Trans. 2 1914, 311. (11) Ayres, G. H.; Booth, M.

H.J . Am. Chem. Soc. 1955, 77, 825.

oxidation of esters,2a,bwhereas a similar high chemioselectivity was achieved by Sharpless by adding acetonitrile to a conventional RuC13-periodate system.2EA simple explanation apparently points to a specific role played by well-defined complexes of the metal in medium-low oxidation states, likely to be stabilized by good ligands or, at least, by good-cmrdinating solvents. Contrary to a-methylene groups, a-methyl groups of ethersare not affected by the oxidation: thus, the asymmetric methyl n-butyl ether is selectively converted into methyl butyrate (eq 3) and no CH3OCH2CH2CH2CH3 + 2 [ 0 ] CH,OCOCH,CH,CH, + H20 (3)

-

n-butyl formate could be detected. Isopropyl ether was rapidly oxidized to acetone, suggesting a mechanism that involves an initial insertion of the oxygen into the tertiary C-H bond with formation of an hemiketal (eq 4), likely in equilibrium with the corresponding ketone and alcohol (acetone and 2-propano1, eq 5). Most likely, Me2CHOCHMe2+ [O] Me2C(OH)OCHMe2 (4)

-

Me2C(OH)OCHMe2 = Me2C0 + Me2CHOH

(5)

acetone also derives from further oxidation of 2-propanol, as evidenced by the fact that under identical conditions the present catalytic system is capable of effectively converting alcohols into the corresponding carbonyl compound^.^ Quite unexpectedly, secondary ethers bearing methylene groups in @-positions,such as 2,Sdimethyl- and 2-methyltetrahydrofuran, behave differently and do not undergo oxidation at the tertiary carbon atom. The only product detected upon oxygenation of 2,5-dimethyltetrahydrofuran is the @-ketoether 23-dimethyldihydrofuran-3-one, whereas 2-methyltetrahydrofuran yields comparable amounts of y-valerolactone, which is the product of the "normal" oxidation of the a-methylene group, and of 2-methyldihydrofuran-3-one, No 2-methyldihydrofuran-4-one, the other possible 0-keto ether, could be detected (Scheme I). The anomalous and intriguing reactivity of this class of secondary ethers deserves further consideration, since a reactivity order tertiary > secondary > primary is expected in the oxidation of saturated C-H moieties, where loss of hydrogen is the ratelimiting step and substitution at the carbon atom might increase the rate of oxidation. Indeed, this argument, while accounting for the observed lack of reactivity of methyl ethers, does not explain both the poor reactivity of the a-CH groups in 2,Sdimethyl- and 2-methyltetrahydrofuran and the facile oxidation of the /3-CH2 groups close to the tertiary carbon atoms. The fact that oxidation at the P-position is preferred to the attack at the tertiary a-carbon atom strongly suggests that these &positions are also in some way "activated" by the presence of the heteroatom. Indeed, P-ketonization of 2,5-dimethyl- and 2-methyltetrhydrofuran occurs with remarkably high rates, up to ca. 3 turnovers/min (see Table I), thus comparable to the conversions of primary ethers to esters, but much larger than the oxygenation rates of unactivated alkanes,

2978 Inorganic Chemistry, Vol. 29, No. 16, 1990

Bressan e t al.

Table 11. Kinetic Isotope Effects in Ruthenium-Catalyzed Oxidation

of Tetrahydrofuran to y-Butyrolactone" rateb substrate I I11 tetrahydrofuran 42 21 tetrahydrofuran-d, 6.8 4.2 tetrahydrofuran-3,3,4,4-d4 20 10.7 tetrahydrofuran-2,2,5,5-d4

0.1

kHlkD

I

I11 c

6.2

5.0

2. I 2.9Y

2.5Y

1.95

OCatalyst: 2 m M I (c~s-(RuCI,[(CH~)~SO],)) or 2 mM 111 ([RucI(DPP),]PF,). Conditions: substrate, 1 M, in CH2CI,; LiCIO, 0.7 M, in HzO; 22 OC. *From the amount of y-butyrolactone formed. Units: turnovers/h. CValuescalculated from the ratios between total isotope effects (tetrahydrofuran-d,) and the data obtained with tetrahydrofuran-3,3,4,4-d4,.

f

s e

0.0 1 .o

0.0

substrate, M (b)

Scheme I1

0

10

20

catalyst, mM (a)

Figure 2. Conversion rate of 2.5-dimethyltetrahydrofuran to 2,5-dimethyldihydrofuran-3-onewith aqueous LiClO (ca. 0.7 M) catalyzed by [RucI(DPP),]PF,, 2 mM, in CH,CI, at 22 OC; (a) effect of substrate concentration (catalyst, 2 mM); (b) effect of catalyst concentration (substrate, 1 M).

H

by the same catalytic system (e.g., cyclohexane is converted to cyclohexanol and cyclohexanone at a rate of ca. 1 turnover/h under identical condition^).^ The observed "total" isotope effect for perdeuteriotetrahydrofuran is rather large, kH/kD > 5 (Table II), as expected for a hydrogen radical abstraction from an sp3-hybridized C-H bond.I2 The most striking feature is the sizable isotope effect resulting from the deuteration of the 3- and 4-positions only of tetrahydrofuran, kH/kD = 2, which can be hardly attributable to (possibly weak) secondary isotope effects, usually ranging as high as On the contrary, these values are very close to those calculated for the 2- and 5-positions, kH/kD = 2.5-3, thus suggesting a peculiar mechanism for the oxidation of ethers, likely to proceed by a hydrogen atom transfer, involving both a- and @-carbonatoms. The overall mechanistic pathway is depicted in Scheme 11, where oxoruthenium(1V) intermediates are proposed as the species responsible for the oxygen transfer. Although direct evidence of these species has not been obtained in the present investigation, a variety of monooxo derivatives of ruthenium have been reported to oxygenate organic substrates, among which are ethers.I4 On the basis of the information derived from product distribution and kinetic isotope effects, a simultaneous hydrogen abstraction from both a-and @-positionscould be envisaged, similar to a mechanism recently proposed by Meyer for the stoichiometric oxidation of alcohols by oxoruthenium complexes.is The resulting unsaturated ethers may undergo further oxidation, leading to the formation of a carbonyl group located in either the a- or @-position,de( I 2) Mimoun, H.; Saussine, L.;Daire, E.;Postel, M.;Fischer, J.; Weiss, R. J . Am. Chem.Soc. 1983, 105, 3101. (13) March. J . Advanced Orgunic Chemistry, 3rd ed.; Wiley Interscience: New York, 1986; p 197. (14) Yam, V. W. W.; Che, C. M.; Tang, W. T. J . Chem. SOC.,Chem. Cornmun. 1988, 100. (15) Thompson, M. S.; Meyer, T. J. J . Am. Chem. Sor. 1982, 104, 4106.

pending on the availability of hydrogen atoms on each position. Within this reaction scheme not only might the oxidation of ethers exhibit sizable and comparable a- and @-deuteriumkinetic effects, but also, as observed, oxygenation of a-tertiary groups should become unviable. In the case of secondary ethers bearing methyl groups a t the @-position, such as isopropyl ether, the classical mechanism should be operating, with hydrogen abstraction at the a-position only, in view of the larger C-H bond energy of methyl with respect to methylene groups. That primary ethers are selectively oxidized to esters and @-keto ethers are never formed can find an explanation, although sibylline, in the general behavior of enol ethers (the putative reaction intermediates), which seem to undergo ketonization exclusively at the position nearest to the heteroatom, at least with conventional oxometal reagents.I6 In agreement with the above statement, y-butyrolactone was the only oxygenated product found during the oxidation of 2,bdihydrofuran with the present catalytic system, whereas, significantly, 2,5-dimethylfuran was converted into the corresponding 3-keto ether 2,5-dimethylfuran-3-one.However, the oxidations of the above unsaturated ethers are definitely slow (Table I), if compared to the oxygenation rates of the saturated ethers, and this is not fully consistent with the fact that we never observed the expected accumulation of detectable amounts of enol ethers during ether oxidations. Indeed, we also found that the ruthenium-catalyzed reaction between ethers and hypochlorite, which is substantially first-order in the catalyst, shows an unexpected zero-order dependence on the concentration of the substrate at a saturated level (Figure 2), thus possibly suggesting that a relatively stable reaction intermediate between the substrate and the catalyst is formed. The fact that each examined substrate is oxidized at a different rate reasonably rules out the possibility that the rate-determining step is the formation of the active oxometal derivative (in this case the rate should be completely indifferent to the concentration of the substrate). A possible mechanism could be a two-step process involving a prior fast equilibrium between the ruthenium complexes, or their oxo derivatives, and the ethers, followed by the rate-determining hydrogen abstraction from the substrates. Therefore, one can speculate that the ether is never leaving the coordination sphere of the catalyst, by undergoing all the required transformations always coordinated to the metal, likely through the oxygen atom. No evidence for this intermediate could be obtained, since measurements on the reacting mixtures were intrinsically difficult to perform. Thus, whereas the rapid mixing of aqueous hypochlorite with the CHzClzsolutions of the catalysts does not result in any substantial change of the initial characteristic colors of the complexes, i.e. pale yellow for cis-(RuCI,[(CH,),SO],), yellow ( 1 6) Piancatelli, G.;Scettri, A.; DAuria,

M. Tetrahedron Lett. 1977, 3483.

Inorg. Chem. 1990, 29, 2979-2982 for tr~ns-[RuCl~(DPP)~], and red for [RuCI(DPP)~]PF,,quantitative spectral measurements of the organic phases were prevented by the fact that, as soon as stirring was stopped, the CH2C12 solutions turned dark green the change being particularly sharp and fast for ~ ~ ~ - I R u C I ~ [ ( C H , ) ~ S OOn ] , ]stirring, . the organic phases revert again to the initial colors. In some cases, green products could be isolated in the solid state and satisfactorily characterized as oxo-bridged binuclear complexes of ruthenium(llI).I7 The green solutions can be therefore considered a sort of "resting" situation, and indeed, they do not exhibit any absorption in the 780-900-~m-~region, where u ( R u 4 ) are expected to fa11.I8 It is concluded that the following reactions (eqs 6-9)

-+ -

Ru" + CIORuIV=O + C1RuIV=O + RuII Ru1114-RuIII RulV=O

+ CIO-

RuIV=O

S

+ O2+ CIRu" + SO

Ru"

(6) (7)

(8) (9)

are likely to take place when aqueous hypochlorite is allowed to be in contact with the organic phase containing the substrate S and the ruthenium(l1) complexes. When the contact between the two phases is interrupted, the putative oxoruthenium(1V) intermediate is consumed with final formation of oxo-bridged species, (17) Bressan, M.; Morvillo, A. Inorg. Chem. 1989, 28, 950. (18) Marmion, M. E.; Takeuchi, K. J. J . Am. Chem. SOC.1986, 108, 510.

2979

likely catalytically inactive. With excess hypochlorite, the OXO terminal species is re-formed and, as a consequence, oxidations of both the substrate and the oxidant are allowed to continue. All the catalysts examined in this investigation exhibit closely related activities and very similar chemioselectivities, despite the significant differences in the chemical environment and stereochemistry. It is therefore possible that a common and very active species is formed in small amounts by the action of hypochlorite on the various complexes since the onset of the reaction. We believe, however, that the active catalysts should be definite complexes of the metal, since the only common species reasonably deriving from the oxidation of the starting complexes are in fact simple "naked" ruthenium oxides, such as Ru02, and for the latter very different selectivities in the oxygenations of ethers have been reported.2a,b The somewhat higher reactivity exhibited by cis(RUCI~[(CH,)~SOJ,] can be attributed to its solubility in water, where it is present as the aquo derivative ( R u C I ~ [ ( C H ~ ) ~ S O ] ~ (H20)1.l9 Acknowledgment. We thank A. Ravazzolo, CNR, for technical assistance. Financial support to the CNR (Centro di Studio sulla Stabilita e Reattivita del Composti di Coordinazione, Padova, and Progetto Finalizzato Chimica Fine 11) and MURST is acknowledged. (19) (a) Alessio, E.;Mestroni, G.;Nardin, G.;Attia, W. M.; Calligaris, M.; Sava, G.; Zorzet, S.Inorg. Chem. 1988, 27, 4099. (b) Oliver, J. D.; Riley, D. P. Inorg. Chem. 1984, 23, 23.

Contribution from the Department of Chemistry, Kansas State University, Manhattan, Kansas 66506-3701

Molecular Vapor Synthesis: The Use of Titanium Monoxide and Vanadium Monoxide Vapors as Reagents Thomas J. Groshens and Kenneth J. Klabunde* Received September 20, 1989 The production and reactivity of titanium monoxide (TiO) and vanadium monoxide (VO) on a synthetically useful scale were studied by using the techniques and apparatus previously developed for reactions of metal atoms. Cocondensation reactions with chlorine showed that greater amounts of metal atoms were produced as impurities in the vapor when stoichiometric metal monoxides were used as vapor sources than when metal sesquioxides were used. It was found that evaporation of MO and M20,solids from open tungsten boats yielded vapor compositions of M, MO, and M 0 2 similar to those reported earlier from Knudsen cells in the case of vanadium but somewhat different in the case of titanium. The monoxide molecules reacted with chlorine to produce oxometal chlorides in high yields. They reacted with 2,4-pentanedioneand other organics containing relatively acidic hydrogen, to form water as a byproduct via protonation of the oxo moiety; coordination compounds of titanium and vanadium were isolated as acac complexes in high yields.

Introduction In recent years, metal vapor chemistry investigators have explored the use of atomic vapors as synthetic reagents.I2 The same basic techniques and apparatus have been applied to the study of other highly reactive short-lived molecular species including CS? BF? Six2?SiF3,6SO,' and others.' In an attempt to further expand the number of reactive molecular vapor reagents available, ( I ) Klabunde. K. J. Chemistry of Free Atoms and Particles. Academic (2) (3) (4) (5) (6) (7)

Press: New York, 1980. Blackbrow, J. R.; Young, D. Meral Vapor Synthesis in Organometallic Chemistry; Springer: New York, 1979. (a) Moltzen, E.;Klabunde, K. J.; Senning, A. Chem. Reu. 1988, 88, 391-406. (b) Moltzen, E.; Kramer, M.; Senning, A,; Klabunde, K. J. J . Org. Chem. 1987, 52, 1156-1161. Timms, P. L. J . Am. Chem. SOC.1967, 89, 1629. Margrave, J. L.; Wilson, P. W. Arc. Chem. Res. 1971, 4, 145. Bierschenk, T. R.; Guerra, M.A.; Juhlke, T. J.; Larson, S.B.; Lagow, R. J. J . Am. Chem. Soc. 1987, 109,4855. Timms, P. L. Ace. Chem. Res. 1973, 6, 118.

0020-1 669/90/ 1329-2979$02.50/0

this work was undertaken as a systematic investigation of the generation and synthetic usefulness of titanium monoxide (TiO) and vanadium monoxide (VO) molecular vapor. Only two reports have appeared that are concerned with transition-metal oxide vapors as synthetic chemical reagents in low-temperature matrices. Cook and Timmss studied the reactivity of molybdenum and tungsten trioxide vapors, which consist mainly of cyclic trimers along with lesser amounts of dimers, tetramers, and pentamers.' Cocondensation of 2,4-pentanedione with MOO, vapor followed by warming to room temperature produced a green solid that analyzed as M o O ~ ( O ~ C ~ HPresumably ,)~. 1 molar equiv of water was also formed as a byproduct although no mention was made of its presence. With W 0 3 vapor, no reactions occurred with pentanedione if care to avoid ligand pyrolysis was taken. Reactions with methanol, boron trichloride, and other reagents were also described. In general, Cook and Timms concluded that (8) Cook, N. D.; Timms, P. L. J . Chem. SOC.,Dalton Trans. 1983, 239.

0 1990 American Chemical Society